Simon Asplen-Taylor is a chief data officer with multi-sector experience, from his current role in the insurance market at Lloyds via previous berths in leisure, retail, business services and IT. He told DataIQ what makes for a successful CDO and how to embed data science into an organisation.
DataIQ (DIQ): You have been a chief data officer on the client side at Rank and now at Lloyds Insurance. How have your roles differed?
Simon Asplen-Taylor (SA-T): As the CDO at Rank, I had an internal-facing role, looking at the organisation and using data to drive four key things: revenue growth, internal efficiency, reducing risk and customer satisfaction.
At Lloyds, however, what I do is directed at the marketplace, so my role has evolved to be not only about my own organisation, but about 400 participants in the Lloyds marketplace. For this, there is the need to scale outside of your own organisation and ensure that the data strategy is agreed and adopted by the other participants. I am still delivering the same drivers that I was in my internal role, but across a community and driving what they are doing, too.
It is another level of complexity and demonstrates how a data community is critical - you can’t do things to people, you have to do it with them. I have to think about and cater to other organisations’ different timescales and agendas, segmenting markets depending on data needs.
For me, the most challenging aspect is people and their way of thinking. If you think about it, Lloyds has been around for 334 years. It has had the same systems and processes in place for some time, so now as we are embracing digital transformation, it can be hard for people to adjust and to shift mindsets from what they and the industry have always known. It can be a challenge, but seeing change and being able to change a marketplace is rewarding.
DIQ: The role of CDO is still a relatively new one – what do you see it as encompassing?
SA-T: As a CDO, you have to be able to identify where data can help an organisation. Really, you need to start off as an evangelist saying, “this is where it can help.” Then you need to pivot to a more business focus and consider the outcomes - revenue, costs, risk, efficiency, customer and employee satisfaction. So, for the CDO, it’s about viewing through all of these lenses and considering, what else can I do with the data to help deliver these results for the business?
As CDO, you need to own the problem, figure the solution and drive the approach, delivering value on the business objectives. In effect, the CDO is a supporting function to the CEO who, through proof of concept and prototyping, supports business operations by leveraging data strategy and delivering value.
Data is just the truth of your business. The facts around your customer base, your revenue and profits, for example. However, along the way there can be an awful lot of data leakage and that can be a waste of crucial customer information. In overlooking quality and governance of data, it can be a lost opportunity.
Some get confused as to whether the CDO is an IT role or not - it isn’t. It doesn’t matter to whom or where the CDO reports, as long as there’s the focus around business value. The role of the CDO has a wide remit and is ever-evolving, which demands a breadth of experience and broad skillset. An individual in this role must be a leader, and someone that people will listen to in order to understand the significance and value of the data team.
DIQ: You have argued for the importance of data keeping a focus on the business, which requires good communication skills and also a level of political ability. Many CDOs come from technical backgrounds where this hasn’t been necessary - what should they do to prepare themselves?
SA-T: CDOs that have a wholly technical background may struggle with the business side of operations. My advice would be to prepare by getting more involved in the strategy of the business, because effectively what you do in the role of the CDO is aligning with business strategy and helping to drive it. You must be proactive in order to succeed in the role and, often, technologists are reactive to business problems.
In order to prepare, they have to understand and spend time working on business strategy, business operations and people - these are the three constituents of the role. Understanding data and the technology on its own is not enough, the role is much more hybrid and multi-faceted.
DIQ: At DataTick, you probably get to see how different clients are approaching data and analytics. Does this give you an insight into critical success factors?
SA-T: When you’ve worked across many organisations, you get a sense of the success factors for specific industries. You could work for a retailer and understand customer service and customer satisfaction. You could work for an efficient operational organisation and know all there is to know about end-to-end processes. Different organisations in different sectors have different cultures, and the truth is there is no one organisation that is brilliant at everything.
Working with so many different organisations will highlight what works and what fails when it comes to culture, marketing or operational efficiency, for example. Generally, I think you learn more from failures than successes. When something works, you think “great” and move on. When something goes wrong, you analyse the data and investigate it to figure out why. Until you find that fix, you won’t be able to move on and be successful.
DIQ: Data science has enjoyed a huge surge in interest and adoption, but it can also be quite disruptive (including to core CDO areas such as governance). How should a business go about introducing it and what expectations should it have?
SA-T: The operative word here is “science”. Science involves experimentation and data science is no different. Data science is experimental, which is key when it comes to expectations.
When introducing it to a business it’s important to have this mindset - that it might not work out first time or every time. But once processes are tried and tested and something works, the exercise will then go through a level of maturity and can be replicated, which is where you earn the value for the business.
You must get three things from an experiment: firstly, value, because it needs to “do” something; secondly, it needs to help to build out the infrastructure of business to allow for more data science projects, such as introducing a data automation tool like WhereScape to build an agile data environment, for example; and thirdly, learning - what are the key takeaways from the experiment to do, or not do, next time? These are three crucial ingredients in any data science project that will allow a business to continue to learn and improve.
A scientist is a creative who is designing and running experiments. However, if this creative freedom is taken from them and they find themselves in an environment and are told to run the same “successful” piece of code every day, they end up becoming an operational or a support person. It can become a real challenge and generally why data scientists don’t often stay in one role long-term, because their creativity in doing their experiments can be squashed as they are asked to own other activities away from what they know and love - the science.
DIQ: Is putting the right tech stack in place an important part of this?
SA-T: Yes, it is! Once the business, operations and people are aligned, the tech stack is important to deliver value and bring the project together. The right tech stack is crucial for data integration, data management and visualisation, and inevitably embedding the data back into the business systems - your customer relationship management (CRM) system, for example. From the data there needs to be action - this is where the true value lies.
DIQ: Finally, in this highly-disrupted year, have you seen any responses that you hope become a permanent part of how data and analytics are adopted and deployed?
SA-T: This year has been an example of massive change. It has been the ultimate test for the effectiveness of companies’ crisis management, business continuity processes and agility.
From an insight point of view, organisations have realised the need for really fast-moving insights. The ability around process and being able to use data to drive insights quickly in an agile way has never been so sought after before. Responsiveness has been key and for organisations that are not able to respond within days, that’s critical.
Most organisations are also realising that their bricks and mortar businesses are not going to generate much value during a lockdown period and therefore the transition to digital has accelerated, for which, of course, data is a key component. For businesses without those data capabilities, there’s the serious risk of going out of business.
The focus around digital transformation and significance of data I hope will be a permanent response shared across organisations. The widespread impact on businesses has demonstrated the need for businesses to innovate, be resilient and create value from data in order to survive and compete during these challenging times.